Last updated 18 August 2026 · Originally published 20 August 2021
Here’s something almost nobody selling CBD will tell you: the amounts used in the research that made CBD famous are far larger than the amounts in most bottles.
That’s worth understanding — but not for the reason you might expect. It isn’t a licence to take more. It’s a reason to be careful about what a research finding actually tells you.
What the trials tested
A typical retail serving is 10 to 25 mg. Here’s what studies have used:
- Anxiety — 300 to 600 mg. A 2024 phase 3 trial used 300–600 mg daily and met its endpoints,[1] but it tested a synthetic nanodispersible formulation designed for absorption, not ordinary retail oil.
- Epilepsy — 10 to 25 mg per kilogram of body weight, as a prescription medicine under medical supervision with liver monitoring.[14]
- Insomnia — 150 mg nightly in the main trial, which did not beat placebo on most measures.[2]
- Neuropathic pain — one small 2026 trial titrated 38 participants up to 800 mg daily and found benefit.[3] Encouraging, and preliminary.
- Exercise recovery — preliminary studies have tested 17 to 67 mg, with mixed results.[15] The one area where trial doses and retail servings overlap.
A 2019 review of CBD dosing across clinical populations found doses ranging from under 1 mg/kg to 50 mg/kg.[4] That review pooled very different diseases, formulations and routes of administration — it describes what has been studied, and isn’t a dosing guide.
A 10 mg gummy is therefore far below many of the doses tested clinically. That doesn’t mean increasing it would reproduce the trial result.
Why more isn’t simply better
Higher doses haven’t reliably done more. In a hypertension trial using an absorption-enhanced CBD formulation, increasing the daily dose from 225–300 mg to 375–450 mg produced no clear additional reduction in blood pressure.[5] In arthritis, controlled trials have failed at 20–30 mg,[11] at 45 mg[12] and at 600 mg.[13]
There’s a long-standing idea that may explain some of this. In 2015, researchers at the Hebrew University of Jerusalem found that purified CBD produced a bell-shaped dose-response in mice — effective within a window, then tailing off as the dose rose further.[6] When they gave a CBD-rich whole-plant extract instead, the response became dose-dependent.
That study is frequently cited in support of full-spectrum CBD, but it can’t establish that claim. It tested one CBD-rich extract in one mouse model of inflammation. It doesn’t tell us whether full-spectrum products outperform isolate in people, or whether human CBD dosing follows the same curve.
There’s also a safety signal worth knowing about. An FDA-funded trial gave healthy adults 5 mg per kilogram daily — about 350 mg for a 70 kg person — for 28 days. Eight of 151 people taking CBD developed liver enzyme elevations more than three times the upper limit of normal, against none of the 50 on placebo.[7] They resolved on stopping, and nobody developed symptoms. But it’s a real signal at a dose people do take.
Interactions matter too. At 5 mg/kg, CBD raised exposure to the antidepressant citalopram by around 43%.[8]
The bit that isn’t about milligrams
Two things make the number on the label less meaningful than it looks.
Absorption is poor and variable. Swallowed CBD passes through the liver before reaching circulation, and only a fraction may get through — although exactly how much remains uncertain. A systematic review of human pharmacokinetics noted that absolute oral bioavailability has never been reliably determined.[9]
Food changes things substantially. Taking CBD with a fatty meal can increase exposure several-fold compared with an empty stomach.[10] The same capsule, taken differently, can produce a very different level of exposure.
How to approach dose safely
Start with the evidence for your problem. A higher dose doesn’t automatically rescue a treatment that has failed in controlled trials — and trying one without supervision introduces new risks.
Check your medications before starting. CBD can change how the body processes certain drugs, and there’s no universal dose below which interactions become impossible. If you take anything prescribed, ask a pharmacist — see our guide to CBD and drug interactions.
Don’t try to recreate a clinical trial with retail oil. Studies using hundreds of milligrams involved standardised formulations, screened participants and sometimes blood monitoring. Retail products may be absorbed quite differently, and nobody is watching your liver function.
Keep the variables steady. Same product, count the milligrams, take it consistently in relation to food. Decide beforehand what a meaningful improvement would look like — sleeping through, a specific pain level, fewer bad days.
If the labelled dose does nothing, discuss escalation rather than improvising it. At higher doses, adverse effects, interactions and liver-enzyme elevations all become more relevant.
Why products are dosed so low
Worth asking, since the gap is so large.
Retail CBD is sold as a consumer wellness product, not as an indication-specific medicine. Its serving sizes reflect price, regulatory limits and uncertainty about safety as much as evidence of benefit.
Higher-strength products cost more and bring greater concern about adverse effects and interactions. So the dose on the bottle shouldn’t be mistaken for a dose shown to treat a condition — but nor should it be treated as an invitation to exceed the directions.
Where this leaves it
If a typical retail dose did nothing, you can’t conclude that a larger one would work. The research dose tells you what investigators tested — not what you should take.
It also isn’t the only explanation for feeling nothing. Formulation, inaccurate labelling, absorption, meal timing, how consistently you took it, what you were measuring and ordinary symptom fluctuation could all be involved.
The useful lesson isn’t “take more”. It’s that a 10 mg gummy shouldn’t be presented as equivalent to a 300 mg clinical trial — and a 300 mg clinical trial shouldn’t be treated as permission to reproduce that dose at home.
Know the milligrams, the formulation and the evidence for your condition. If the dose starts moving beyond the product directions, involve a clinician or pharmacist rather than experimenting upward.
None of this is medical advice, and prescription CBD should be dosed by a clinician.
For more, see our guide to what the research actually shows, or browse Using CBD.
References
- Gundugurti PR, Banda N, Yadlapalli SSR, et al. Evaluation of the efficacy, safety, and pharmacokinetics of nanodispersible cannabidiol oral solution versus placebo in mild to moderate anxiety subjects. Asian Journal of Psychiatry. 2024;97:104073. doi:10.1016/j.ajp.2024.104073
- Narayan AJ, Downey LA, Rose S, Di Natale L, Hayley AC. Cannabidiol for moderate-severe insomnia: a randomized controlled pilot trial of 150 mg of nightly dosing. Journal of Clinical Sleep Medicine. 2024;20(5):753–763. doi:10.5664/jcsm.10998
- Robertson RV, Suraev A, McCartney D, et al. High-dose cannabidiol for chronic neuropathic pain associated with spinal cord injury: a randomised clinical trial. eClinicalMedicine. 2026;96:103986. doi:10.1016/j.eclinm.2026.103986
- Millar SA, Stone NL, Bellman ZD, Yates AS, England TJ, O’Sullivan SE. A systematic review of cannabidiol dosing in clinical populations. British Journal of Clinical Pharmacology. 2019;85(9):1888–1900. doi:10.1111/bcp.14038
- Dujic G, Kumric M, Vrdoljak J, Dujic Z, Bozic J. Chronic effects of oral cannabidiol delivery on 24-h ambulatory blood pressure in patients with hypertension (HYPER-H21-4). Cannabis and Cannabinoid Research. 2024;9(4):979–989. doi:10.1089/can.2022.0320
- Gallily R, Yekhtin Z, Hanuš LO. Overcoming the bell-shaped dose-response of cannabidiol by using cannabis extract enriched in cannabidiol. Pharmacology & Pharmacy. 2015;6(2):75–85. doi:10.4236/pp.2015.62010
- Florian J, Salcedo P, Burkhart K, et al. Cannabidiol and liver enzyme level elevations in healthy adults: a randomized clinical trial. JAMA Internal Medicine. 2025;185(9):1070–1078. doi:10.1001/jamainternmed.2025.2366
- Salcedo P, Volpe DA, Chaturbedi A, et al. Clinical study to evaluate drug interactions of cannabidiol with citalopram and morphine in healthy adults. Clinical Pharmacology & Therapeutics. 2026;119(4):1095–1104. doi:10.1002/cpt.70219
- Millar SA, Stone NL, Yates AS, O’Sullivan SE. A systematic review on the pharmacokinetics of cannabidiol in humans. Frontiers in Pharmacology. 2018;9:1365. doi:10.3389/fphar.2018.01365
- Taylor L, Gidal B, Blakey G, et al. A phase I, randomized, double-blind, placebo-controlled, single ascending dose, multiple dose, and food effect trial of highly purified cannabidiol in healthy subjects. CNS Drugs. 2018;32(11):1053–1067. doi:10.1007/s40263-018-0578-5
- Vela J, Dreyer L, Petersen KK, et al. Cannabidiol treatment in hand osteoarthritis and psoriatic arthritis: a randomized, double-blind, placebo-controlled trial. Pain. 2022;163(6):1206–1214. doi:10.1097/j.pain.0000000000002466
- Mojoli A, Haider O, Fakih Y, et al. Effects and safety of a CBD-rich Cannabis sativa oil in knee osteoarthritis (CANOA). Frontiers in Pharmacology. 2025;16:1657065. doi:10.3389/fphar.2025.1657065
- Pramhas S, Thalhammer T, Terner S, et al. Oral cannabidiol as add-on to paracetamol for painful chronic osteoarthritis of the knee. The Lancet Regional Health – Europe. 2023;35:100777. doi:10.1016/j.lanepe.2023.100777
- Epidiolex (cannabidiol) oral solution: US prescribing information, revised May 2026.
- Stauffer JW, Crow JA, Bishop MD, Cook RL, Borsa PA. Efficacy and safety of cannabidiol on reducing pain and functional impairment associated with exercise-induced muscle injury. Journal of Cannabis Research. 2026;8:64. doi:10.1186/s42238-026-00431-x
About this article
Substantially rewritten in August 2026. This update sets out the gap between the doses used in research and those in retail products, and why that gap is a reason for caution rather than an argument for taking more. With earlier reporting by Lance Griffin.
